(Revised) DESCRIPTION: (Adapted from Applicant's Abstract) Relationships between alcohol and other drug use and risky behavior will be studied using cross-sectional data from approximately 600 adolescents with substance use disorders (SUDs). This study will focus on event-specific relationships between substance use and risky behavior where both occur during the same occasion. Analyses will address three conclusions about these relationships from research: (1) they are too complex to be analyzable; (2) they can be explained by underlying or predisposing third factors; and (3) they are mediated by the decision-making process of weighing the costs and benefits of unprotected encounters. The specific aims of the study are to: 1. Determine the correlation's between different types of substance use and risky behavior of varying levels of perceived risk. 2. Test three third-factor explanations for observed relationships between substance use and risky behavior. 3. Test whether components of decision-making mediate relationships between third factors and risky behavior or between substance use and risky behavior. Approximately 250 SUD subjects, as well as 350 community controls, will come from the Pittsburgh Adolescent Alcohol Research Center (PAARC). The protocol for these subjects has been expanded to include detailed questions about behavior. Subjects will receive one assessment and one 6-week period of phone-in diary interviews. Results from this study can be used to help design and target HIV prevention for adolescents by (1) suggesting realistic program content about the effects of alcohol and drugs, as well as indicating who is most susceptible to these effects, (2) discovering risk factors and behaviors that are unique to adolescent females, and (3) providing information about factors that affect decision-making in.